White blood cell (WBC) scintigraphy is considered the nuclear medicine imaging gold standard for diagnosing osteomyelitis in the diabetic foot. Recent papers have suggested that the use of (18)F-FDG PET/CT produces similar diagnostic accuracy, but clear interpretation criteria have not yet been established. Our aim was to evaluate the role of sequential (18)F-FDG PET/CT in patients with a high suspicion of osteomyelitis to define objective interpretation criteria to be compared with WBC scintigraphy.
Methods: Thirteen patients whom clinicians considered positive for osteomyelitis (7 with ulcers, 6 with exposed bone) were enrolled. The patients underwent (99m)Tc-exametazime WBC scintigraphy with acquisition times of 30 min, 3 h, and 20 h and sequential (18)F-FDG PET/CT with acquisition times of 10 min, 1 h, and 2 h. A biopsy or tissue culture was performed for final diagnosis. Several interpretation criteria (qualitative and quantitative) were tested.
Results: At final biopsy, 7 patients had osteomyelitis, 2 had soft-tissue infection without osteomyelitis, and 4 had no infection. The best interpretation criterion for osteomyelitis with WBC scintigraphy was a target-to-background (T/B) ratio greater than 2.0 at 20 h and increasing with time. A T/B ratio greater than 2.0 at 20 h but stable or decreasing with time was suggestive of soft-tissue infection. A T/B ratio of no more than 2.0 at 20 h excluded an infection. Thus, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for osteomyelitis were 86%, 100%, 100%, 86%, and 92%, respectively. For (18)F-FDG PET/CT, the best interpretation criterion for osteomyelitis was a maximal standardized uptake value (SUVmax) greater than 2.0 at 1 and 2 h and increasing with time. A SUVmax greater than 2.0 after 1 and 2 h but stable or decreasing with time was suggestive of a soft-tissue infection. An SUVmax less than 2.0 excluded an infection. (18)F-FDG PET at 10 min was not useful. Using these criteria, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for osteomyelitis were 43%, 67%, 60%, 50%, and 54%, respectively. Combining visual assessment of PET at 1 h and CT was best for differentiating between osteomyelitis and soft-tissue infection, with a diagnostic accuracy of 62%.
Conclusion: (18)F-FDG PET/CT, even with sequential imaging, has a low diagnostic accuracy for osteomyelitis and cannot replace WBC scintigraphy in patients with diabetic foot.